Research On The Blind Estimation Of The Parameters Of Continuous Phase Modulated Signals

The need for information transmission is growing,while spectrum resources are becoming increasingly scarce.Therefore,new modulation methods should be developed to meet communication needs.Continuous Phase Modulation(CPM)is a nonlinear memory modulation.Because of its continuous phase,constant envelope,high spectral efficiency and high efficiency,it is widely used in modern communication systems.In the non cooperative communication system,the modulation and the modulation parameters of sender are unknown to the receiver.Thus,the modulation parameter estimation must be applied to get parameter information of the sender,in order to realize the reception and blind demodulation of transmitted signal.However,it is difficult to estimate CPM signal,especially for multi-h CPM,due to its multiple modulation parameters.The modulation index cyclic variation makes the structure of CPM signal more complex than the single index CPM signal,and the blind estimation is more difficult.Thus,this thesis studies the correlation modulation parameters and SNR blind estimation of CPM signal.Modulation parameters of CPM mainly include carrier frequency,symbol period,modulation exponent,modulation order and pulse phase correlation length.Firstly,the mutual dependence between the carrier frequencies and the symbol periodic blind estimation are mutually dependent,because of that,the joint blind estimation of CPM signal carrier frequency and symbol period can be finished.The estimation accuracy of carrier frequency is 0.01,and the symbol period achieves 100% correct idebtification probability,when the SNR reaches 10 d B.Secondly,the cycle stability of CPM signal is analyzed.It is found that when the modulation exponent is a singular integer,spectral lines appear in the cyclic spectra.And the number of spectral lines is related to the modulation order.The blind estimation of the single exponential CPM signal modulation exponent and the modulation order are finished,with the help of estimation algorithm based on the cyclic stability.The modulation index of the multi-h CPM signal can not be estimated directly,because it is cyclic.In this paper,the modulation index of the same code element is collected into a group,forming multiple single exponential CPM signals.Then estimating the modulation index using the above mentioned method realizes the estimation of multiple modulation exponents for full response multi-h CPM signals.What's more,it is found that the pulse phase correlation length affects the instantaneous frequency of the signal,so the pulse phase correlation length is accurately identified using SVM algorithm,combined with the estimation of the above parameters.Finally,this study conducts blind estimation of the SNR of single-h and multi-h CPM signals,based on higher-order moments and signal subspace decomposition algorithms.The applicability of the algorithms is verified through simulation.In the end,the applicable range of the two algorithms is given.Using the MATLAB simulation tool,this thesis simulates and analyzes the single-h CPM signal and multi-h CPM signal based on the simulation model constructed in this study.The results show that the proposed algorithm has a more accurate estimation than the existing blind indexing algorithm with modulation index,symbol period,and carrier frequency offset.This algorithm reduces the computation cost to a certain extent,and shortens the estimation time.It is easy to realize in engineering project.Also,the algorithm is applicable to the blind estimation of multi-h CPM signal modulation parameters.